Evaluating complex interventions involves determining both whether and how they work; such evaluations are theoretically and practically complex. First, the components of care themselves may interact positively or negatively with each other and also may be required in different doses or formats depending on context. The kind of interventions required to bring about change are also likely to be multi-faceted; educational, audit-based and facilitation-based interventions each have evidence to support them. The choice and application of appropriate outcome measures at different levels of change is challenging. Lastly, implementation often occurs within the shifting sands of health services reform and development.
Realistic Evaluation is a relatively new framework for understanding how and why interventions work in the real world and has been recommended as a means to understand the dissemination of service innovation. Analysis focuses on uncovering key mechanisms and on the interactions between mechanism and context in order to develop ‘middle range theories’ about how they lead to outcomes. Accumulation of evidence, which may be qualitative or quantitative, and may also be derived from external sources, leads to a refining of these theories. Realistic Evaluation, therefore, promises to be a useful framework for understanding the key functions of an intervention by examining its relationship with the context.
Shared care emphasizes the need for co-ordination between primary care and specialists to reduce duplication and address unmet need; chronic disease management focuses on service redesign in primary care incorporating timely review, expert input, patient involvement and information systems.